On Tue, Jan 16, 2018 at 9:33 PM,  <ahwhit...@gmail.com> wrote:
> I am connecting 2 bb blacks via serial uart, /dev/ttyO1. or /dev/ttyS1. they
> are symlinked. I need to use them at the fastest speeds, i.e. B3000000 or
> more. When i use the older 7.11 image (straight off beagleboard downloads
> site), the BB's communicate rock solid, no losses. When i upgrade the image
> to the current 9.1, the non-IoT image, and use the same code and the same
> HW, i can't get reliable data xfer'd across the serial link. A lot of errors
> in comparison to the almost none using the old image. It finally was working
> reliably at B500000.
>
> My dilemma is i don't want to remain with effectively an obsolete image
> (i.e.. debian 7.11). It makes everything else harder to keep up. I've played
> with uEnv.txt options. Nothing seems to matter.
>
> I didn't find any thread in the forum discussing this kind of problem.
> Clearly something is different about 9.1 and the serial subsystem from the
> earlier 7.11 version. I was having same problem with latest version 8 as
> well.
>
> has anyone experienced this, and better, found a fix. Driver, device tree
> configuration, whatever.

You can install the old "3.8.x" based kernel in 9.x or 8.x via:

cd /opt/scripts/tools/
git pull
sudo ./update_kernel.sh --bone-channel --stable
sudo reboot

BTW, newer images us U-Boot Overlays, so remember to double check
/boot/uEnv.txt and "disable it":

from:
enable_uboot_overlays=1

to:
#enable_uboot_overlays=1


Then you can verify if it's a kernel or new lib issue..

Regards,

-- 
Robert Nelson
https://rcn-ee.com/

-- 
For more options, visit http://beagleboard.org/discuss
--- 
You received this message because you are subscribed to the Google Groups 
"BeagleBoard" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to beagleboard+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/beagleboard/CAOCHtYjFoU8wbJC2o%2B58N3yXNTXhtgRgV6RuYuEUnRxfE2DqQA%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.

Reply via email to